Emergency systems for power failures are known from many fields of engineering. In particular, in the field of data-processing, it is necessary to prevent data losses due to uncontrolled disconnections of data-processing devices as a result of power failures and current interruptions. Suitable emergency systems should in these cases temporarily take over the supply of power to at least the most important components of data-processing devices to enable a proper shutdown.
The safeguarding of volatile memories of data-processing devices both with rechargeable batteries and with capacitors is known. A rechargeable nickel-metal hydride battery which is basically suitable for applications of this type is described, for example, in EP 1 011 163 B1. Suitable capacitors are, in particular, so-called “double-layer” capacitors (supercapacitors). The use of the latter as energy source for an emergency system for power failures is described, for example, in DE 20 2004 017 545 U1.
Double-layer capacitors have the great advantage that they can very rapidly supply very high pulsed currents. However, the capacity of those capacitors is limited according to the nature of a capacitor. Furthermore, like all capacitors, they have a falling voltage characteristic.
Rechargeable batteries such as those described, for example, in EP 1 011 163 B1 offer a significantly higher capacity. However, these have the disadvantage that they can be overcharged at the relatively high temperatures that can occur in data-processing devices during operation, which can lead to a drastic reduction in their life expectancy.
There is, therefore, a need to provide an emergency system for power failures which does not have the above disadvantages.